SYNTHESIS REPORT: PRE-INTERVENTION ASSESSMENTS OF PRIMARY HEALTH CARE AND PREVENTION SERVICES AT THE COMMUNE LEVEL IN KHANH HOA AND DA NANG PROVINCES

Transcription

1 SYNTHESIS REPORT: PRE-INTERVENTION ASSESSMENTS OF PRIMARY HEALTH CARE AND PREVENTION SERVICES AT THE COMMUNE LEVEL IN KHANH HOA AND DA NANG PROVINCES Report Prepared for Atlantic Philanthropies by the Population Council 2 Dang Dung Street Hanoi, Vietnam June 30, 2005

2 ACKNOWLEDGMENTS The Population Council expresses its sincere thanks to Atlantic Philanthropies for inviting the Council to conduct pre-intervention assessments of primary health care and prevention services at the commune level in Khanh Hoa and Da Nang provinces. We are very grateful to the leaders and senior staff of Khanh Hoa and Da Nang Health Services for their close and effective collaboration in the assessments. We offer our sincere thanks to the staff of two provincial health services, including special mention of returned fellows of Population Council who served as active and enthusiastic members of the assessment team, especially in analysis of existing data. In addition, we are also indebted to health leaders and administrators, public and private health and non-health professionals from selected districts and communes/wards for their collaboration in participation into focus group discussions and in-depth interviews, and providing us with valuable information on the situation of primary health care in these two provinces. We also thank the commune women and men who are enthusiastic participating in the focus group discussions and in-depth interviews that provided knowledge and insights on residents' health seeking behaviors the in community. Finally, we appreciated our consulting colleagues from Hanoi Medical University and other national research institutions in Hanoi who collaborated with us in field work, data analysis, and writing of reports of these assessments. The authors 2

3 TABLE OF CONTENTS Executive Summary... 4 I. INTRODUCTION II. PROJECT OBJECTIVES AND APPROACH III. DATA FOR THIS REPORT ) The Types of Data ) Preparation of Data Sets V. FINDINGS ) Human resources, training, and service delivery policies ) Infrastructure, equipment, and supplies ) Operation of curative and some key national preventive programs ) Factors affecting the demand for CHC services VI. NEXT STEPS: CHOICE AND TESTING OF INTERVENTIONS TO IMPROVE THE SYSTEM OF CARE

4 Executive Summary The government health system of the Socialist Republic of Viet Nam (GOVN) is wellrespected outside Viet Nam, in particular for its achievements in preventive health and disease control. However there is evidence that the primary level of care, based at the commune health center (CHC), falls short in meeting the actual and perceived needs of many users and would-be users. This is more often the case for those who are poor, living in remote areas, or ethnic minorities. Current GOVN priorities and policies have begun to make conditions more favorable for intensified attention to primary health care (PHC) and public health at commune level. Therefore, it is timely to increase the levels of good evidence, data and analysis of primary health care and prevention services, in order to better diagnose issues and problems and to test feasible solutions. In response to the present context for primary health care policies and services in Viet Nam, the Provincial Health Services of Da Nang and Khanh Hoa reached an agreement in 2004 with the Population Council to conduct coordinated assessments of primary health care services in each province. Since 1989, the Population Council, an international nongovernmental organization, has worked with Vietnamese partners to assess and improve quality and access of health services at the grass roots level. This project, which is supported by Atlantic Philanthropies, aims to add new, practical knowledge about the specific opportunities and challenges that Khanh Hoa and Da Nang Provincial Health Services will face as they formulate plans for improving primary health care in the public health system. This executive summary distills the methods, experience and findings of this joint project. In two of its sections (one on intervention ideas emerging from the assessments, and the other on recommended modifications in the assessment method for use in future provinces), this summary presents some emerging views and ideas that have not yet been fully discussed with the Khanh Hoa and Da Nang partners. Thus these sections should be read as more preliminary than the earlier sections. Brief Overview of Da Nang and Khanh Hoa Provinces. In 2001: Da Nang has average GDP growth rate: 12.2% per annum, and GDP per head: VND 7.9 million and Khanh Hoa has average GDP growth rate: 9% per annum, and GDP per head: VND 6.1 million (source: MPI webiste), As profiled in the 2003 Statistical and Health Statistical Yearbooks, Khanh Hoa and Da Nang are quite different. Khanh Hoa is located in south-central Viet Nam, has a population of 1,110,000, of whom 80% live in rural areas. Its geography is large and varied, from mountains in the west to a long seacoast in the east. Nha Trang, the capital city, is a growing tourist and resort center. Health indicators for 2003 included 52 major obstetrical complications with 2 deaths; a malaria rate of 422 per 100,000, and a death rate of 0.27; TB cases totaled 1,159; and cumulative HIV/AIDS cases were 1,168 and 407 deaths. One in four children under the age of five was underweight. Road accidents numbered 355 including 247 deaths. Residents of Khanh Hoa s rural mountainous regions, many of poor or near-poor status, have difficult access to health services. The public health system in Khanh Hoa includes five Provincial hospitals, and 133 commune 4

5 health centers (CHCs) for 137 communes. 64% of CHCs have medical doctors, and 97% have midwives. Child immunization coverage was very high, at 98% of children under one year of age. Da Nang, in central Viet Nam, is a city-province which ranks as the fourth largest urban area in the nation. Its expanding population stood at 747,000 in Only 12 % of the province is rural. In recent years, economic growth has shown a steady rise, while poverty has steadily declined. Indicators for some important health problems in 2003 included 24% of children under the age of five who were underweight; major obstetric complications numbered 69, including 2 deaths; and the malaria rate was 26 per 100,000 with no deaths. TB prevalence stood at 1,339 cases, and cumulative HIV/AIDS cases were 747. Road accidents totaled 216 including 147 deaths. The public health system includes 8 provincial hospitals, and 47 CHCs, or one CHC in every commune. 75% of CHCs have doctors, and 100% have a midwife. Almost all children one year of age and under have been vaccinated for measles and other diseases. Project approach This study aimed to: Develop assessments in 2 provinces, in order to inform and guide Atlantic s support of a future Commune Health Services Initiative in a cross-section of strategic provinces. Use project team composition and methods that insure that strong local leadership and external experts work together at each stage of the project. Engage the Provincial Health Service in each selected province as the co-director of the assessment, along with Population Council Viet Nam Office. Document and assess preventive and primary health care services at commune level from three angles o community, i.e. users, residents, and leaders o commune-based health providers and managers, and o health officials at province and district levels Develop and test an assessment methodology for each province that makes systematic use of existing databases and studies before doing limited collection of new data in each province. Aim for an efficient core methodology that can be used in other provinces if found useful in the initial two. These objectives have all proved to be appropriate and achievable, as is documented in three detailed written final products: individual reports for each province, and a synthesis report for which this is the executive summary. Both provinces chose to design their individual assessments according to the same basic structure and methods. Each assessment was organized to assess current services in terms of four sets of factors that are key determinants of access and quality of services. Three sets of factors are on the supply side: 1. policies and human resources, 2. infrastructure, equipment and medicines, and 3. Functions and practice of preventive and curative services, and management practice. The fourth set of factors is on the demand 5

6 side: perceptions of local residents and community leaders/officials, and health seeking behaviors of residents. The framework and scope of the assessment both warrant comment in terms of their relative novelty. The incorporation of in-depth collection of demand-side, communitylevel data is rare in pre-intervention baseline studies designed or commissioned by governments. More distinctive still is the scope, which is the whole span of primary health care and prevention services. The norm is still to use the single issue or single problem organizing principle for health reform initiatives what is usually called the vertical program approach. Methods and data In all, twenty CHCs from each province were selected for the appraisal, following a purposeful sampling strategy designed to balance the need for statistically representative quantitative data with the need for a wide range of qualitative data from community residents and leaders, as well as from CHC managers and staff, and their district and province level supervisors. Use of existing data. Qualitative analysis was supplemented with statistical analysis of existing health information system (HMIS) data extracted from the routine monthly and other periodic reports that each CHC is required to prepare using standard forms issued by the national Ministry of Health (MOH). These data from the 40 study CHCs spanned the 22-month period beginning in April, All HMIS data were computerized and analyzed for consistency and data quality. Six caseload indicators were found to be consistently reported. Although it turned out that the small sample size, reporting gaps, and wide monthly variations combined to prevent cross-sectional analyses, the project team found that time trends were robust, which provided the basis for an exploratory analysis of the determinants of caseload trends. Collection of limited new data in the field. In December 2004, the field researchers visited 40 CHCs in Khanh Hoa and Da Nang, to compile more detailed data on equipment and supplies available at CHCs, and on worker and user perceptions of the system of care. In a sub-set of 12 communes, 8 in Khanh Hoa and 4 in Da Nang, quantitative and qualitative data were compiled through focus group discussions (FGDs) and in-depth interviews (IDIs) with community residents and leaders, and village health workers. Research teams also conducted FGDs and IDIs with health leaders and managers, and a small sample of private health workers. Key findings 1) Policies and human resources In the two provinces, primary health services are respected social assets that local residents view as bringing multiple benefits to every commune and most villages. 6

7 When focus groups were convened, discussion of curative health services predominated, despite the agreed importance of preventive care to the success of the Viet Nam health program. Access to physicians. Most CHCs now have physicians on staff, which is highly appreciated by residents and commune leaders. Where physicians are not part of the regular CHC staff, the local population views this as a major deficiency in the quality of care. Access to paramedics. Other types and numbers of staff at CHCs are generally considered sufficient by local residents, and are consistent with MOH national guidelines. While physicians are often discussed, local residents were less interested to discuss paramedical staff. Access to village health workers (VHWs) at community level. VHW networks were established in most rural and mountainous communes as an important network to support CHCs at community level -- in primary care, health education, and detection and response to new health problems. However, the VHW system is available in rural areas only. In urban areas, such as Da Nang city, VHWs are not used. Instead, there are health collaborators, individuals who do some of the same work. In Da Nang many said the collaborators needed more training, and viewed VHWs as better-trained and thus more effective. For their part, VHWs said they had little or no technical training, and lack essential equipment, instruments, and pharmaceuticals. Low or no compensation for their work is the rule, and provincial authorities state that budget funds from the center for support of these workers is lacking. Health insurance and equity. Since 2001, all CHCs have been required to provide services for patients at the commune/ward level under the health insurance policy. This policy has been expanded to the poor since 2002, to create a health care fund for the poor, which is viewed very positively, especially by poor residents of the two provinces. Health insurance for people in general and for the poor in particular is important in helping reduce their health expenses. However the proportion of the population covered by insurance in most communes is still quite modest. This, combined with the rapid growth of wealth in the middle class in Viet Nam, creates documented increases in disparities in access. This issue was not a special focus for the two provinces assessments, but growing disparities is a widely recognized trend in Viet Nam, and is acknowledged with concern by the government. The range and quality of curative services. Interviews with health managers indicated that some types of career advancement are well developed for CHC staff. This was corroborated by reviewing records, which showed that a substantial number of CHC staff have participated in advanced training in curative care. Most notable is that assistant physicians can be sent for one-year training that qualifies them to become physicians. One province has more than 60 assistant physicians who are nearing completion of such training. Also, general 7

8 assistant physicians can take an advanced course that qualifies them to add a specialty in obstetric and gynecological care. Despite these advances in extending the range of services, the assessment teams review of data, probing and further discussion identified various problems: o Equipment gaps and shortcomings. None of the CHCs visited in this appraisal fully conformed to national benchmarks for equipment in primary health care facilities. In addition, existing equipment too often is not in proper working order. In general, CHCs are equipped for paramedical services rather than medical services, which often prevent physicians from performing diagnostic and treatment duties that are commensurate with their training and skills. o o In-service training. While the staffing patterns and qualifications of staff meet basic needs, in-service training is not always well organized or designed for ensuring that staff can meet increasing or changing health care needs. Health staff reported attending short term training courses as available in such topics as health program management, reproductive health, child health care, TB control, leprosy, and HIV/AIDS. Rising demand for other curative services. Community residents and leaders expressed considerable demand for additional types of specialist services. The cost of adding a second physician to CHC staff is unlikely to be affordable. Other models, for example visiting specialists, may merit review. 2) Infrastructure, equipment and pharmaceutical supplies Facility layout. In general, there are adequate essential conditions available for examination and treatment at most CHCs. Basic equipment and supplies. Most CHCs have basic equipment and medicines that are needed for essential services. However emergency equipment and supplies were not adequate even at a basic level. Construction and major renovation. Quantitative analysis of HMIS data showed that construction or major renovation of a CHC has a slight, but significant, positive effect on how much commune residents use CHC services. If this small effect is projected over time, and assumed to continue, the long-term effect can be substantial. Thus, small effects on monthly caseloads may be important in the long run. Qualitative research results corroborate this conclusion, suggesting that user perceptions of the quality of care are affected by the quality of the facility. Nonetheless, although construction and major renovation showed an impact, other factors explained most of the variation in statistical models. 8

9 Construction and distance. In general, CHCs are located near roads and the mean distance from the CHC to the district and provincial level health providers was not great ranging from 2km in the cities to 35 km in Khanh Hoa rural and mountainous districts. However, evidence from this appraisal demonstrates that construction has a greater impact on caseloads as distance increases. This suggests that investment in construction or major renovation in remote CHCs will have a greater impact than construction in CHCs that are relatively close to referral points. Utilities. Most CHCs had adequate electrical supply for the operation of machines, for treatment and emergency treatment, and for daily activities. Nearly all CHCs in this appraisal had a telephone or other communication capabilities. Land area. Land areas of CHCs were large enough for construction and implementation of primary health care and prevention services and expansion in the future. 3) Functions and practice of preventive and curative services, and management practice. Preventive health services. While the preventive health service system was less commonly discussed by focus group participants than curative care, preventive service activities are a major, and often dominant, component of CHC staff work routines. This focus is widely appreciated for its achievements in such areas as near-universal child immunization coverage, compliance with safe delivery practice guidelines, and high levels of contraceptive use. However, organizational problems were often noted in discussions, especially by CHC managers and workers. In particular, vertical programs typically have their own independent targets and goals, and reporting requirements and forms. CHC staff must coordinate the requirements of an average of about 20 national health programs, each with unique reporting and task requirements. Programs are sometimes promulgated in top-down, target-driven directives that are inconsistent with local priorities, needs, and capabilities. There is a need for further investigation of work routines required for an integrated regimen of primary health care services that addresses general needs and priorities, yet also provides flexibility for CHC staff to adapt strategies and activities to local circumstances. Curative health services. CHC managers and staff reported that there were not enough physicians and health staff to adequately meet the demand for CHC services. Administrative matters and meetings occupy a major component of medical staff time. This is viewed by CHC staff and clientele as a factor constraining access to medical services. Discussants often praised the quality of CHC staff, indicating respect for their dedication to quality health services. Nonetheless, perceptions of shortcomings in quality often detract from demand for CHC services: In particular, discussants noted inadequacies in counseling, privacy and referrals; infection control; and deficient equipment repair and 9

10 maintenance. Discussions suggest that perceptions of the quality of CHC care are a dominant factor constraining demand for CHC services. More and better management information. There is a pressing need to simplify HMIS operations and improve data feedback and use so that data management operations can serve the monitoring needs of frontline workers and clientele. This project designed a set of tools for this purpose which proved feasible for regular use by existing staff. If these tools were instituted by provincial health service teams, findings could be integrated into management decision-making processes rather than standing apart in a technical report. Such reports would be regular management tools and not require special research. 4) The Demand-side: Discussion of factors affecting health seeking behaviors. Qualitative discussions and interviews involving local residents and community leaders and officials provided many insights into the health seeking behaviors of residents and their rising expectations of CHC services. Access to primary health care. CHCs are crucial sources of access to primary health care for local patients, particularly in the rural and mountainous areas. The geographic characteristics of the coastal and mountainous areas of Khanh Hoa and rural areas of Da Nang constrain CHC activities and affect efforts to seek and provide services. The climate of care and demand. In general, local residents report being satisfied with services provided by CHCs, and particularly with the friendly attitude of health staff, which contributes to patient s confidence and reliance on the CHCs. When the public perceives that CHCs have poor infrastructure, inadequate equipment, or limited supplies, residents report lowers their use of CHCs as primary service points. Insurance. The national health insurance scheme plays a crucial role in shaping demand for CHC services among the poor. Medical examination and treatment for the poor have reduced the burden of medical costs for clients, particularly for ethnic minority groups and other poor residents of the mountainous and other rural areas. Private care. General curative care caseloads of CHC in this appraisal are gradually declining with time, in part because clientele have resources for private services that they view as having higher quality than the CHC alternative. Development of private health services helps reduces the overload of work for public health centers at higher levels, creating more choices for clients in seeking health services relevant with their capability and condition. Nonetheless, there is a need to review the policy implications for investment in CHC services of the growing reliance on private care. Health-seeking behaviors. Residents and users employ a wide range of health care options: 10

11 o o o o Residents report routine use of self-treatment and private providers, often in consultation with pharmacies, sometimes simultaneous with use of the public system, and sometimes as alternatives to public system. Residents also report selective use of the public system, taking minor problems to CHCs while pursuing self-referral to DHC or provincial facilities. Per capita utilization of CHCs is higher in rural and remote areas than in urban areas. While residents are generally positive about CHC staff and services, they seek a wider range of specialized services than CHCs can realistically provide, including services such as ear, nose and throat specialists, ultrasound exams, x- rays, dental care and eye care. 11

12 I. INTRODUCTION The government health system of the Socialist Republic of Viet Nam (GOVN) is wellrespected outside Viet Nam, in particular for its achievements in preventive health and disease control. However there is evidence that the primary level of care, based at the commune health center (CHC), falls short in meeting the actual and perceived needs of many users and would-be users. This is more often the case for those who are poor, living in remote areas, or ethnic minorities. Current GOVN priorities and policies have begun to make conditions more favorable for intensified attention to primary health care (PHC) and public health at commune level. It is timely therefore to increase the levels of good evidence, data and analysis of primary health care and prevention services, in order to better diagnose issues and problems and to test feasible solutions. In response to the present context for primary health care policies and services in Viet Nam, the Provincial Health Services of Da Nang and Khanh Hoa reached an agreement in 2004 with the Population Council to conduct coordinated assessments of primary health care services in each province. Since 1989, the Population Council, an international nongovernmental organization, has worked with Vietnamese partners to assess and improve quality and access of health services at the grass roots level. This project, which is supported by Atlantic Philanthropies, aims to add new, practical knowledge about the specific opportunities and challenges that Khanh Hoa and Da Nang Provincial Health Services will face as they formulate plans for improving primary health care in the public health system. Brief Overview of Da Nang and Khanh Hoa Provinces. In 2001: Da Nang has average GDP growth rate: 12.2% per annum, and GDP per head: VND 7.9 million and Khanh Hoa has average GDP growth rate: 9% per annum, and GDP per head: VND 6.1 million (source MPI webiste), As profiled in the 2003 Statistical and Health Statistical Yearbooks, Khanh Hoa and Da Nang are quite different. Khanh Hoa is located in south-central Viet Nam, has a population of 1,110,000, of whom 80% live in rural areas. Its geography is large and varied, from mountains in the west to a long seacoast in the east. Nha Trang, the capital city, is a growing tourist and resort center in Viet Nam. Health indicators for 2003 included 52 major obstetrical complications with 2 deaths; a malaria rate of 422 per 100,000, and a death rate of 0.27; TB cases totaled 1,159; and cumulative HIV/AIDS cases were 1,168 and 407 deaths. One in four children under the age of five was underweight. Road accidents numbered 355 including 247 deaths. Residents of Khanh Hoa s rural mountainous regions, many of poor or near-poor status, have difficult access to health services. The public health system in Khanh Hoa includes five Provincial hospitals, and 133 commune health centers (CHCs) for 137 communes. 64% of CHCs have medical doctors, and 97% have midwives. Child immunization coverage was very high, at 98% of children under one year of age. Da Nang, in central Viet Nam, is a city-province which ranks as the fourth largest urban area in the nation. Its expanding population stood at 747,000 in Only 12 % of the province is rural. In recent years, economic growth has shown a steady rise, while poverty has steadily declined. Indicators for some important health problems in 2003 included 24% of children under the age of five who were underweight; major obstetric 12

13 complications numbered 69, including 2 deaths; and the malaria rate was 26 per 100,000 with no deaths. TB prevalence stood at 1,339 cases, and cumulative HIV/AIDS cases were 747. Road accidents totaled 216 including 147 deaths. The public health system includes 8 provincial hospitals, and 47 CHCs, or one CHC in every commune. 75% of CHCs have doctors, and 100% have a midwife. Almost all children one year of age and under have been vaccinated for measles and other diseases. II. PROJECT OBJECTIVES AND APPROACH The Population Council and the Provincial Health Services of Khanh Hoa and Da Nang reached an agreement in the fall of 2004 to prepare an assessment of primary health care. This project, which was supported by Atlantic Philanthropies, aimed to add new, practical knowledge about the specific opportunities and challenges that Provincial Health Services will face as they formulate plans for improving primary health care in the public health system. This study aimed to: Develop assessments in 2 provinces, in order to inform and guide Atlantic s support of a future Commune Health Services Initiative in a cross-section of strategic provinces. Use project team composition and methods that insure that strong local leadership and external experts work together at each stage of the project. Engage the Provincial Health Service in each selected province as the co-director of the assessment, along with Population Council Viet Nam Office. Document and assess preventive and primary health care services at commune level from three angles o community, i.e. users, residents, and leaders o commune-based health providers and managers, and o health officials at province and district levels Develop and test an assessment methodology for each province that makes systematic use of existing databases and studies before doing collection of new data in each province. Identify and test a methodology for pre-intervention assessment that can be used in other provinces if found useful in the initial two. These objectives have all proved to be appropriate and achievable. Both provinces chose to design their individual assessments according to the same basic structure and methods. Each assessment was organized to assess current services in terms of four sets of factors that are key determinants of access and quality of services. Three sets of factors are on the supply side: 1. policies and human resources, 2. infrastructure, equipment and medicines, and 3. Functions and practice of preventive and curative services, and management practice. The fourth set of factors is on the demand side: perceptions of local residents and community leaders/officials, and health seeking behaviors of residents. The framework and scope of the assessment both warrant comment in terms of their relative novelty. The incorporation of in-depth collection of demand-side, communitylevel data is rare in pre-intervention baseline studies designed or commissioned by 13

14 governments. More distinctive still is the scope, which is the whole span of primary health care and prevention services. The norm is still to use the single issue or single problem organizing principle for health reform initiatives what is usually called the vertical program approach. III. DATA FOR THIS REPORT 1) The Types of Data This report for Khanh Hoa Province and Da Nang is based on three sets of existing and new data, as outlined in the next section on methodology: Data Set A, Existing Quantitative data: from MOH-required routine reports (sometimes referred to as the Health Management Information System, or HMIS); Data Set B, New Quantitative data: Survey of commune health center-based health providers and Inventory of CHC amount and status of staff positions, infrastructure, equipment, supplies, etc.; and Data Set C, New Qualitative field data: In-depth interviews and Focus group discussions with commune residents, commune and health leaders, village health workers, and local private health providers. This data set also includes in-depth interviews with health system leaders from provincial and district levels. Since some, but not all, communes in Khanh Hoa and Da Nang have been engaged in health systems improvements, these two provinces represent an important resource for assessing whether improvements affect: 1) demand for services as measured by patient or client caseload, or 2) the quality of care, as indicated by qualitative data recording the views of clients and health service providers. All findings from Data Set A are based on the combination of data from Da Nang and Khanh Hoa Provinces, so that analyses are conducted on statistically robust sample sizes. We also present findings from Data Sets B and C delineated by the source of information, Khanh Hoa or Da Nang, as well as the type of respondent. Taken together, findings identify key factors determining the quality, accessibility, and volume of primary health care services. The report paints a picture of many parts, with no single factor that dominates. It offers new evidence, both confirming and challenging some widely accepted priorities for primary health care, and raises topics for consideration by the two provinces as they plan initiatives to improve primary health care. 2) Preparation of Data Sets Selection of communes: A nested purposeful sample. In all, twenty CHCs from each province were selected for the appraisal following a purposeful sampling strategy. Selection was designed to address competing goals: the appraisal requires large-scale, statistically representative data; yet, a wide range of quantitative and qualitative data from community members, CHC staff, and district medical personnel are needed to provide 14

15 insights into health seeking behavior and service delivery practices. New quantitative and qualitative information required for these goals could not be collected on a large scale; yet large-scale information was needed to provide the statistical power for inference. Therefore, a combined nested approach was employed to build on the advantages of each analytical strategy, while avoiding intractable data collection designs. Figure 1 illustrates this nested design study. In each province, communes for Group 1 & 2 were selected in pairs: First, we chose 10 communes that have new CHCs in each province: Group 1 was prepared of CHCs that have been renovated during the period from Dec to Dec Communes were eliminated from the list with construction in 2004 so that CHCs included in the sample had populations served by renovated CHCs or not served by new CHCs. Then, districts were arranged by population size so that in the largest districts 2 communes were selected and the smallest, one district, leading to the random sampling of 10 communes. For each new CHC we selected a matched old CHC. These were purposefully sampled as the nearest CHC that matched the corresponding new CHC. Where there were CHC of equivalent distance, the matched CHC was the oldest eligible CHC. Figure 1: The Design Of The Nested Purposeful Sample, Groups I, II, And III. Group III= 40 CHC Existing data only: Data Set A Group II = 20 CHC Inventory & SA staff + existing data (Data sets A&B) Group I=12 CHC C Q ualitative + Inventory & SA + Existing data 15

16 Group 1 was comprised of twelve communes -- eight in Khanh Hoa and four in Da Nang -- where all elements of the study were conducted. The data Group I data collection package included: Compilation of existing archival data from the routine MOH data reporting forms, as translated and presented in English in Appendix B. Compilation of new quantitative data based on visits to CHCs to conduct a detailed a facility inventory and interviews with two staff at each CHC using a structured questionnaire. New qualitative data collected from in-depth research recording CHC staff views on health service issues; community residents experience and views on health seeking behavior and service options; provincial and district health managers/leaders views on policies and challenges for local CHCs and prevention services; and Group 2 was comprised of 20 communes--12 communes in Khanh Hoa, (including the 8 which are also in Group 1), and 16 in Da Nang (including the four in Group 1). This yields 20 CHCs where quantitative elements of the study were conducted, including situation analysis inventories and staff interviews. This level of data collection ensured feasibility, avoiding the need for expensive and complex requirements of in-depth data collection and interpretation activities. Group 3 would ideally be comprised of all the remaining CHCs in each study province. However, owing to time and resource constraints, existing CHC Health Information System Data were entered for the 20 CHCs in Group 1 and 2 only and 10 additional CHCs in each province, making 40 altogether. Data for the analysis are extracted from existing records. All CHCs in Vietnam are required to submit monthly reports of total caseloads of preventive and curative care to district health authorities. These standard forms are reproduced in Appendix B. To provide data for the present analysis, these monthly reports have been computerized for 40 study CHCs in Da Nang and Khanh Hoa Provinces, providing time trend data from the period when forms were introduced, April, In Khanh Hoa, some CHCs report data for 22 months (from March 2002 to December 2003) while others reported less than 22 times over this period. Da Nang CHCs reported data for 21 months (from April 2002 to Dec 2003). Available reports from the 40 Data Group III CHC shave been merged so that observation time is scored for the 22 months from March 2002 to December Taken together these Data Set A reports define a set of time series data for monthly clinic caseloads, including indicators defining the completion month of construction for the 20 study CHCs where facilities were upgraded. Owing to the high frequency of missing entries on some variables on some of the records, analysis focuses on eight indicators of the volume of health services that were consistently reported by most participating CHCs: Total curative and preventive visits (an indicator of the total volume of care) 16

17 Total in-patients and out-patients (an indicator of the total demand for curative care) Gynecological examinations Gynecological treatment cases All diarrheal disease cases Under age five diarrheal disease cases All pneumonia cases Under age five pneumonia cases The analysis of Data Set A proceeded in three stages: First, we reviewed general time trends for each indicator to determine if caseloads are increasing, decreasing, or remaining the same. Next, we reviewed time trends in reference to the timing of construction or renovation to determine if the post construction period is associated with caseload levels that are higher than preconstruction levels or if levels of caseloads change in response to construction. Finally, we examine the effect of construction on caseloads, adjusting for other possible determinants. This final analysis explains the variation in caseload levels over time. Dataset A limitations. Various factors point to a need for caution in interpreting specific results from the statistical analysis that follows: Problems arise from the small sample used for this study. Results are highly sensitive to outliers. For example, observations from one CHC include exceptionally high caseloads. This CHC was renovated in June 2001, but the actual caseload was affected by the Expanded Immunization Program. During this campaign many children were immunized and if any child was found with respiratory symptoms such as cough or fever, ARI medication was prescribed and the child was reported as a pneumonia case. Construction and renovation are timed and located according to administrate standards that aim to solve problems. This selective placement of facility investment may have biased results in unmeasured ways. Data quality of CHC HMIS operations was flawed. In the absence of routine use of HMIS data by management omissions, errors, and inconsistencies in how the routine data are recorded go unnoticed. Potentially useful indicators of health operations and outputs could not be used for this assessment because data quality was too poor for the type of analysis to be pursued. Time trend data compensate for flaws, since assumptions can be made about trends that are unbiased by factors that affect data quality and a particular point in time. Nonetheless, routine use of HMIS data for management purpose, with tools for reporting findings and errors back to CHC, would address these problems and widen the scope for analyses. 17

18 Collection of new data at field level: Data Sets B and C. Collection of new data took place in 20 Ward/Commune health centers and communities in each province. For Khanh Hoa province, these communes/wards were randomly selected from 137 communes and wards of Khanh Hoa province, among which 16 communes are rural and 4 are urban wards These were selected by a communes/ward stratification procedure that aimed to generate a sample that is geographically typical of the province, including 10 rural communes, six mountainous, one coastal commune and three communes with mixed terrain. All CHCs selected for the study were accessible by staff of a higher level health facility. Nearly all of the sample CHC were accessible by tarred roads and were functioning as centers that refer cases to district hospitals. The average distance from the CHC to the regional polyclinic is two kilometres, to district hospital is 5.2 km, and to the provincial/city hopital is 35.4 km. Thus, in comparison with other regions throughout the country, the average distance from the sample CHCs in Khanh Hoa province to higher levels of referral is rather short-- results from the National Health Survey 2001/ show that the average distance from CHCs to district hospitals is 12 km (10 km in the plain areas and 22 km in the mountainous areas), and from CHCs to the provincial hospital it is 51 km (34 in the plain areas and 88 km in the mountainous areas). The surveyed CHCs are located in the communes with the typical economies of Khanh Hoa province - agriculture, aquaculture and service. The communities under this survey were rather developed with telephone coverage and electric network services in most of the surveyed communes; in the rural areas, deep wells were available almost everywhere; piped water was available in the urban areas and there were either deep wells or natural water in the mountainous areas. Focus group and in-depth interviews were convened of the following study participants: Health managers and leaders of Khanh Hoa province: In all, 34 Health managers from provincial, 8 district and 8 CHCs participated in in-depth interviews. Most of them are aged 40 and the majority are male (85%). They had different duration of working experience, from one month to 20 years. Health sector work experience of the interviewees varied between years. Most of them are physicsians (70%), among them, one-third are general physicians, the others are physicians of other specilialities, mainly gynaecology, odonto-stomatology, epidemiology and public health. One-third of the participants completed a graduate education (Specialist I) equivalent to a masters degree in the medical field. Health staff: A total of 59 health staff from 20 CHS participated in interviews with structured questionnaires. The average age of health staff is 36 and 80% of them are married. Female health staff account for 66%, and male staff, 40%. Most of them (93%) are Kinh majority. Regarding professional levels: 7% are general physicians, 2% are physicians of different specialities and 17% are obstestric and gynaecologic assistant physicians, 32% are assistant physicians of other specialities; 20% are secondary and primary midwives; 19% secondary and 1 Ministry of Health - General Statistic Office - Report on the National Health Survey : State Primary Health Services Pages

19 primary nurses. The average time of working in the health sector is 13 years; their average time of working in current health centers is 10 years. Private health providers: A total of 16 private health providers in eight communes participated in in-depth interviews. They are of different ages, 43.6 years old on average. The oldest health provider is 86 years old while the youngest is 28. More than two- thirds of the participants are male. Regarding their professional background, four of them are physicians, four are traditional healers, five are pharmacists and three are assistant physicians. Most of them had a long working experience with 15.6 years of working in the health sector. Professional experience ranges between two years to 43 years. Most are experienced working in the public health system. At the moment, more than one-third of them are still working in public health facilities such as commune health services, hospital or pharmaceutical companies. One-fourth of the private health providers did not have experience working in any public health facilities. Among interviewed health providers, one-third provide western health services, nearly one-third are providing traditional health services and another one-third operate a pharmacy. Apart from one participantwho provided western health care services without a legal permit due to insufficient number of years of experience, all other health providers hadpermits to run their medical operation according to their trained professional background. Village health workers (VHW): 54 VHWs participated in FGD in 8 communes, of which over 90 percent are female. They are of middle age on average (43.4 years old). Over half of all VHWs pursue agricultural livelihoods, and have 3.5 years of experience in this role. Community leaders: 77 community leaders (including commune and village leaders, members of women unions and population boards, etc.) participated in eight FGDs about primary health care and preventive medicine in the commune health centers. The average age of leaders was 45 years old, and they are mainly male of Kinh majority and married with 3 children on average. They mainly do agricultural production, accounting for 61%. Nearly 42% of them were working as village heads or vice heads. The rest pursue other careers as members of women unions or youth associations. Residents: In Khanh Hoa, 100 individuals participated in 8 FGDs and 37 in-depth interviews in 8 communes. In Khanh Hoa, respondents are mainly Kinh majority (75%), and Raglay ethnic minority (24%). A very small percentage is Tay minority. Most of the interviewees are engaged in agricultural production (64%), or private business (27%). Some are retired workers. For Da Nang Province, collection of new data in the field also took place in 20 Ward/Commune health centers, of which six are commune health centers in the rural communes and 14 are ward health centers in the urban area. As a big city in the center of the country, Da Nang s geographical characteristics include terrain that is 45% plains, 30% coastal, 10% midlands, 10% mountainous and 5% mixed. 19

20 Among the 20 surveyed CHCs in Da Nang city, the average distance from the CHCs to district hospitals is 5.2 km, and to provincial hospitals is 9.7 km. Thus, the average distance between the CHC to district and to provincial hospitals is not far compared with that of other CHCs in the country. Results from the National Health Survey 2001/ show that the average distance from CHCs to district hospitals is 12 km (10 km in the plain areas and 22 km in the mountainous areas), from CHCs to provincial hospitals was 51 km (34 in the plain areas and 88 km in the mountainous areas). There are tarred roads leading to all the 20 surveyed CHCs in Da Nang, so accessibility is convenient. Focus group and in-depth interviews were convened of the following study participants: Health leaders of Da Nang City: 19 Health Leaders from provincial and 4 districts and managers of 4 CHCs participated in in-depth interviews. The average age was 54.7, and the majority are male (95%). Work experience ranged from one to 20 years. 68.4% have more than 20 years of work experience in the health sector. Most are physicians (95%), among them, one-third are general physicians, the others are physicians of other specilialities, mainly gynaecology, odontostomatology-maxillo Facial Surgery (OSM), epidemiology and public health. 50% of them attended graduate training (for doctor with first level of specialization). Health staff: 60 health staff from 20 CHS participated in interviews with structured questionnaires. The average age of health staff is 38 years. Female staff accounted for 80% and are all King majority. Regarding professionalism: 15% are general physicians, 10% are physicians of different specialities and 11.7% are obstestric and gynaecologic assistant physicians, 32% are assistant physicians of other specilities; 25% are secondary midwives and 3% are primary midwives; 2% primary nurses (no secondary nurses), 2% have other specialities, there are no traditional healers. Private health providers: 11 private health providers in 4 communes participated in in-depth interviews. Their average age was older than other participants, at 53.2 years; the oldest health provider is 64 years old while the youngest is 43. More than half of them are male. Regarding professional level, two are physicians, three are traditional healers, four are secondary pharmacists, and one midwife. The combined average of work experience is 23 years in the health sector, ranging between ten and 34 years. Most of them have worked in the public health system. At the moment, more than half of them are still working in the public health facilities such as commune health services, hospitals or pharmaceutical companys, including three agents for pharmaceutical companies and two pharmacists. Among the interviewed health providers, four out of 11 provide western health services, three provide traditional health services and four operate pharmacies. Apart from one nurse who was not granted a legal permit, all other health providers hold permits to run medical operations according to their trained profession. 2 Ministry of Health - General Statistic Office - Report on the National Health Survey : State Primary Health Services Pages